Diagnostic usefulness of percutaneous liver biopsy in HIV-infected patients with fever of unknown origin

• Published in: The Journal of infection 1999-03, Vol.38 (2), p.94-98
• Main Authors: García-Ordóñez, M.A., Colmenero, J.D., Jiménez-Oñate, F., Martos, F., Martínez, J., Juárez, C.
• Format: Article
• Language: English
• Subjects: Adult; Biological and medical sciences; Biopsy, Needle - methods; CD4 Lymphocyte Count; Female; Fever of Unknown Origin - etiology; HIV - pathogenicity; HIV Infections - complications; Homosexuality; Human immunodeficiency virus 1; Human viral diseases; Humans; Infectious diseases; Leishmaniasis, Visceral - complications; Leishmaniasis, Visceral - diagnosis; Leishmaniasis, Visceral - pathology; Liver - microbiology; Liver - pathology; Liver Diseases - complications; Liver Diseases - diagnosis; Male; Medical sciences; Multivariate Analysis; Predictive Value of Tests; Retrospective Studies; Risk Factors; Spain; Substance Abuse, Intravenous; Tuberculosis - complications; Tuberculosis - diagnosis; Tuberculosis - pathology; Viral diseases; Viral diseases of the lymphoid tissue and the blood. Aids
• ISSN: 0163-4453; 1532-2742
• DOI: 10.1016/S0163-4453(99)90075-0

Objectives: to determine the value of percutaneous liver biopsy (PLB) in the diagnosis of fever of unknown origin (FUO) in HIV-infected patients and establish a prediction model for its usefulness to enable diagnosis of FUO in these patients to be standardized. Methods: a total of 58 HIV-infected patients who underwent PLB for the evaluation of FUO were studied at ‘Carlos Haya’ Hospital in Malaga, Spain. The patients were classified into three groups, according to the results of the PLB: (a) diagnostic PLB (when a definitive diagnosis was obtained); (b) helpful PLB (the tissue sample showed suggestive, but not definitive, findings); and (c) normal or non-specific PLB (no contribution to diagnosis, the findings being normal or irrelevant). Multivariate analysis was made to establish a prediction model for the diagnostic usefulness of PLB, calculating the positive (PPV) and negative (NPV) predictive values. Results: PLB was carried out in 58 HIV-infected patients during diagnosis of FUO. Risk factors for HIV infection included intravenous drug use (72.4%), homosexual or bisexual activities (12.1%), and heterosexual transmission (15.5%). Fifty-two out of 58 patients (89.6%) had previous AIDS-defining illnesses. The mean CD4 lymphocyte count ± SD was 56.4 ± 90.9/mm 3. The mean duration of fever was 43 days. Diagnosis could established in 51 (87.9%) patients, with tuberculosis (50%) and leishmaniasis (20.7%) being the most common. The PLB was diagnostic in 25 cases (43.1%), helpful in 13 (22.4%), and normal or non-specific in the remaining 20 (34.5%). Biopsy-associated complications occurred in two cases. The presence of hepatomegaly or splenomegaly were the most useful factors in predicting the usefulness of the PLB, with a PPV of 86.1% and NPV of 68.2%. In patients with tuberculosis, an increased alkaline phosphatase and hepatomegaly had a PPV of 86.4% and a NPV of 71.4%. Conclusions: PLB is a useful technique for the diagnosis of FUO in HIV-infected persons. Early PLB should be considered in those patients with hepatosplenomegaly and increased alkaline phosphatase levels.